# google-research/fixmatch

A simple method to perform semi-supervised learning with limited data.

Repository: https://github.com/google-research/fixmatch
Canonical: https://ross.abutalabs.com/products/fixmatch
Language: Python
License: Apache-2.0
License Family: permissive
Archived: true
Last push: 2024-07-30T20:49:39+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 0, release rhythm 8, longevity 100
- inputs: {"age_days": 2416, "days_push": 764, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: archived
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1223, forks 190 (observed 2026-08-28T04:04:02.687946+00:00)

## What it is
Official research code for FixMatch, a semi-supervised learning method that combines consistency regularization with confidence-based pseudo-labeling to train image classifiers with very few labeled examples. It includes dataset preparation scripts and training pipelines for CIFAR, SVHN, STL-10, and ImageNet experiments.

## Use cases
- train an image classifier with only a handful of labeled examples
- reproduce the FixMatch paper results
- run semi-supervised learning experiments on CIFAR-10 or SVHN
- compare SSL methods under low-label regimes
- adapt consistency-based pseudo-labeling to a new dataset

## When to choose
- you need state-of-the-art semi-supervised image classification with limited labels
- you want the reference implementation of the FixMatch algorithm
- you are doing SSL research and need a solid baseline

## When to avoid
- you need a production-ready or maintained ML library
- you work outside image classification
- you want a simple high-level API rather than research scripts

## Facets
- artifact type: library
- maturity: maintenance
- function: machine-learning, deep-learning, llm-training
- domain: machine-learning, deep-learning
- platform: python
- tags: semi-supervised-learning, research-code, tensorflow, paper-reproduction, image-classification, research, gpu, linux

## Member repositories
- google-research/fixmatch (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:02.687946+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T06:15:32.610597+00:00, confidence not recorded.
  - readme: https://github.com/google-research/fixmatch (fetched 2026-08-28T04:04:02.687946+00:00, sha 8637de251854)
- Data as of 2026-08-30T08:39:29.467469+00:00.